cinara

Appearance

Speech ​

Speech turns text into lifelike speech. Pick a voice, a language and an emotion, and download the audio.

The page ​

Speech fills the screen: the script on the left, and a panel on the right with two tabs, Settings and History. On a laptop the page itself doesn't scroll; the script and the panel do.

Make speech ​

  1. Open Speech in the sidebar, or choose Generate speech on Home.
  2. Type or paste your text. With the box empty, Get started with offers ready scripts: Narrate a story, Voice a trailer, Record an ad, Introduce a podcast, Guide a meditation, Teach a lesson, Play a character and Speak Hindi. Each one fills in the text and picks a voice that suits it.
  3. In Settings, choose a Voice and a Model, and change anything else you need.
  4. Choose Generate speech under the text, or press ⌘ Enter (Ctrl Enter on Windows). The price shows next to it.

The audio plays under the script when it's ready. Use the download button on the player to save it.

Settings ​

Voice ​

The chosen voice shows as one row; the play button beside it plays a preview. Choose the row to open every voice:

  • Search by name or description.
  • Filter by All, Female or Male, by My voices, or by what a voice suits: Narration, Conversational, Characters, Ads & trailers, Education, Social media, Kids.
  • Choose the play button on any voice to hear it, and the voice itself to pick it.
  • Your workspace's own voices (cloned and designed) come first, marked yours.
  • Cinara voices have a name that matches the language, for example "Mentor · Nathan" in English and "Mentor · Raghav" for Hindi. Names for picks which names to show: this script (automatic) follows your language or text, or choose a region (Global English, India, Arabic, Spanish, Portuguese, French, German, Chinese, Japanese, Korean, Indonesian & Malay). Previews speak in that region's language.
  • Browse the Voice Library opens the Voice Library.

Model ​

  • Speech: 1 credit per character.
  • Speech HD: 2 credits per character.

Language ​

Detect, or choose one of 37 languages: English, Hindi, Spanish, French, Portuguese, German, Arabic, Russian, Japanese, Korean, Chinese, Cantonese, Italian, Turkish, Dutch, Ukrainian, Vietnamese, Indonesian, Thai, Polish, Romanian, Greek, Czech, Finnish, Bulgarian, Danish, Hebrew, Malay, Slovak, Swedish, Croatian, Hungarian, Norwegian, Slovenian, Catalan, Nynorsk and Afrikaans.

Emotion ​

Match the text (the default), Neutral, Happy, Sad, Angry, Fearful, Disgusted or Surprised.

Speed and pitch ​

  • Speed: 0.50× (slower) to 2.00× (faster).
  • Pitch: −12 (lower) to +12 (higher).
  • Reset to defaults appears once you've changed a setting.

Download format ​

  • MP3 · 128 kbps (default)
  • MP3 · 256 kbps · 44.1 kHz
  • FLAC · lossless · 44.1 kHz: up to 4,000 characters. For longer text, choose MP3.

Pauses and sounds ​

  • Add a pause with a tag like <#0.5#> (a half-second pause).
  • Add a reaction with Add a sound under the text. It puts a tag at your cursor: Laughs, Sighs, Coughs, Clears throat, Gasps, Sniffs, Groans or Yawns. You can also type them, for example (laughs).
  • Subtitles leave these tags out.

Pronunciations ​

Teach Cinara to say a word differently. Rules apply to Speech, Dialogue and Studio audiobooks across your workspace.

  1. In Settings, choose Pronunciations.
  2. Type the Word, and how to Say it as.
  3. Choose Add.

To remove a rule, choose the × next to it. Close the window with × or Esc. Rules match whole words and ignore capital letters.

History ​

The History tab lists your last 30 speech generations, newest first, with the voice, length, credits and when.

  • Play each one in its small player, or download it.
  • Use settings: copy its text and settings back into the script and Settings.
  • SRT subtitles and VTT subtitles: download subtitles timed to the audio. The first request takes a moment (Making subtitles…).
  • A failed item shows Failed · refunded.

Limits ​

  • Up to 50,000 characters per generation. The counter under the text shows how many you've used.
  • Text over 4,000 characters is made in parts and joined into one file. The bar under the text shows "N parts, joined into one file".
  • FLAC: up to 4,000 characters.
  • Pronunciations: the word up to 100 characters, the replacement up to 200. Each word can have one rule.

Credits ​

  • Speech: 1 credit per character.
  • Speech HD: 2 credits per character.
  • The minimum is 1 credit.
  • Credits count the text as you wrote it, not after pronunciation rules change it.
  • Subtitles don't cost extra credits.
  • If generation fails, your credits are refunded.

Tips ​

  • Set the Language when you know it. It helps pronunciation and picks the right voice names.
  • Try Speech for drafts, then Speech HD for the final take.
  • Use pronunciations for brand names, people's names and jargon.
  • Use Use settings to tweak a take without retyping.

Cinara · AI voice, music and video

© 2026 Cinara